mlsanityData to 14 Oct 2026

Fair NLP Models with Differentially Private Text Encoders

arXiv 2205.06135 ↗First posted 12 May 2022cs.CL cs.LGcited by 9

Not on the map. Why?

Abstract (arXiv)

Encoded text representations often capture sensitive attributes about individuals (e.g., race or gender), which raise privacy concerns and can make downstream models unfair to certain groups. In this work, we propose FEDERATE, an approach that combines ideas from differential privacy and adversarial training to learn private text representations which also induces fairer models. We empirically evaluate the trade-off between the privacy of the representations and the fairness and accuracy of the downstream model on four NLP datasets. Our results show that FEDERATE consistently improves upon previous methods, and thus suggest that privacy and fairness can positively reinforce each other.

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 42 of its 65 references that are in the map. First posted on arXiv 12 May 2022.

Builds on

42 of 65 references.

References · 42

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  3. 2014Deep Learning Face Attributes in the Wildcited by 3,693
  4. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,281
  5. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,917
  6. 2016Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddingscited by 1,312
  7. 2018Gender Bias in Coreference Resolution: Evaluation and Debiasing Methodscited by 659
  8. 2020Null It Out: Guarding Protected Attributes by Iterative Nullspace Projectioncited by 455
  9. 2018Learning Adversarially Fair and Transferable Representationscited by 353
  10. 2019Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Settingcited by 349
  11. 2019Differential Privacy Has Disparate Impact on Model Accuracycited by 316
  12. 2014RAPPOR: Randomized Aggregatable Privacy-Preserving Ordinal Responsecited by 272
30 more
  1. 2019Lipstick on a Pig: Debiasing Methods Cover up Systematic Gender Biases in Word Embeddings But do not Remove Themcited by 268
  2. 2018Examining Gender and Race Bias in Two Hundred Sentiment Analysis Systemscited by 234
  3. 2020Information-Theoretic Probing with Minimum Description Lengthcited by 232
  4. 2020Information Leakage in Embedding Modelscited by 203
  5. 2016Demographic Dialectal Variation in Social Media: A Case Study of African-American Englishcited by 196
  6. 2019Mitigating Bias in Algorithmic Hiring: Evaluating Claims and Practicescited by 174
  7. 2018Adversarial Removal of Demographic Attributes from Text Datacited by 170
  8. 2017Data Decisions and Theoretical Implications when Adversarially Learning Fair Representationscited by 165
  9. 2017Collecting Telemetry Data Privatelycited by 149
  10. 2017Fader Networks: Manipulating Images by Sliding Attributescited by 118
  11. 2017Using millions of emoji occurrences to learn any-domain representations for detecting sentiment, emotion and sarcasmcited by 118
  12. 2022Linear Adversarial Concept Erasurecited by 101
  13. 2020On the Privacy Risks of Algorithmic Fairnesscited by 82
  14. 2018Differentially Private Fair Learningcited by 76
  15. 2018Towards Robust and Privacy-preserving Text Representationscited by 75
  16. 2020Differentially Private Representation for NLP: Formal Guarantee and An Empirical Study on Privacy and Fairnesscited by 67
  17. 2019Fair Decision Making using Privacy-Protected Datacited by 58
  18. 2018Privacy-preserving Neural Representations of Textcited by 54
  19. 2019What's in a Name? Reducing Bias in Bios without Access to Protected Attributescited by 51
  20. 2021Diverse Adversaries for Mitigating Bias in Trainingcited by 48
  21. 2020Double-Hard Debias: Tailoring Word Embeddings for Gender Bias Mitigationcited by 36
  22. 2019Disparate Vulnerability to Membership Inference Attackscited by 35
  23. 2021ADePT: Auto-encoder based Differentially Private Text Transformationcited by 31
  24. 2021When differential privacy meets NLP: The devil is in the detailcited by 25
  25. 2021CAPE: Context-Aware Private Embeddings for Private Language Learningcited by 22
  26. 2019Conceptor Debiasing of Word Representations Evaluated on WEATcited by 18
  27. 2015Building a RAPPOR with the Unknown: Privacy-Preserving Learning of Associations and Data Dictionariescited by 16
  28. 2021Adversarial Scrubbing of Demographic Information for Text Classificationcited by 14
  29. 2020Removing Disparate Impact of Differentially Private Stochastic Gradient Descent on Model Accuracycited by 13
  30. 2018To Reverse the Gradient or Not: An Empirical Comparison of Adversarial and Multi-task Learning in Speech Recognitioncited by 10

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 2; 2023: 3; 2024: 2; 2025: 1; 2026: 1.

9 papers.

  1. 2022Debiasing Methods for Fairer Neural Models in Vision and Language Research: A Surveycited by 22
  2. 2023Differentially Private Natural Language Models: Recent Advances and Future Directionscited by 20
  3. 2024A Comparative Analysis of Word-Level Metric Differential Privacy: Benchmarking The Privacy-Utility Trade-offcited by 13
  4. 2022FairGrad: Fairness Aware Gradient Descentcited by 8
  5. 2023Holistic Survey of Privacy and Fairness in Machine Learningcited by 4
  6. 2023Fair Without Leveling Down: A New Intersectional Fairness Definitioncited by 1
  7. 2024Differentially Private Multimodal Laplacian Dropout (DP-MLD) for EEG Representative Learningcited by 1
  8. 2025Investigating User Perspectives on Differentially Private Text Privatizationcited by 1
  9. 2026PrivBench: A Holistic and Modular Benchmarking Platform for Evaluating Text-to-Text Privatizationcited by 0